run-llama/liteparse · error

failed to extract pdfium archive

Error message

failed to extract pdfium archive

What it means

After creating the temp dir, download_pdfium unpacks the downloaded tar/zip archive into it with the `tar` crate's unpack, panicking with this message on failure. Failures here mean the archive is corrupt, truncated, or in an unsupported format (e.g. a failed download that saved an HTML error page, or a platform archive variant the extractor can't read).

Solutions

1. Delete the cached/partial download in the pdfium cache dir and rebuild to force a fresh download.
2. Check network/proxy: ensure the pdfium release URL is reachable and not intercepted (curl the URL and inspect the response).
3. Verify the archive manually (tar -tf / unzip -l) to confirm integrity and format.
4. Retry the build after a transient network failure; if behind a proxy, configure HTTPS_PROXY properly.

Example fix

# before: corrupt cached archive
$ cargo build
   ... panicked: failed to extract pdfium archive

# after: purge cache and retry
$ rm -rf ~/.cache/liteparse/pdfium*
$ unset HTTP_PROXY HTTPS_PROXY  # or fix proxy config
$ cargo build

# Verify the archive downloads intact before building
curl -fSL -o /tmp/pdfium.tar.gz "$PDFIUM_URL" && tar -tzf /tmp/pdfium.tar.gz > /dev/null || { echo "bad download"; exit 1; }

# Build wrapper: retry once after clearing a failed extraction
for i in 1 2; do
  cargo build && break
  rm -rf ~/.cache/liteparse/pdfium*
done

When it happens

Trigger: Running `cargo build` where pdfium-sys downloads a pdfium release archive and unpack fails: network proxy/captive portal returned an HTML error page saved as the archive, the download was truncated, the tar contains unsupported entry types (symlinks/hardlinks on Windows), or the archive format doesn't match the expected one.

Common situations: Corporate proxies intercepting the download URL; flaky networks mid-download; GitHub releases changing their archive layout; restrictive sandboxes that strip symlink entries from the tarball.

Related errors

• failed to create temp dir
• USERPROFILE env var not set
• HOME env var not set
• failed to move pdfium to cache dir
• WASI proc_exit called with code

AI-assisted analysis of run-llama/liteparse@22d2dd8cd7 (2026-09-08). Data as JSON: /api/errors/c7ad83d64f67936a. Report an issue: GitHub.

Appendix: source

Thrown at crates/pdfium-sys/build.rs:188

    eprintln!("pdfium-sys: GET {url}");

    let response = ureq::get(&url).call().unwrap_or_else(|e| {
        panic!("failed to download pdfium from {url}: {e}");
    });

    let reader = response.into_body().into_reader();
    let gz = flate2::read::GzDecoder::new(reader);
    let mut archive = tar::Archive::new(gz);

    // Extract to a temp dir first, then rename atomically
    let tmp = dest.with_extension("tmp");
    if tmp.exists() {
        fs::remove_dir_all(&tmp).ok();
    }
    fs::create_dir_all(&tmp).expect("failed to create temp dir");
    archive
        .unpack(&tmp)
        .expect("failed to extract pdfium archive");

    // Fix dylib install name on macOS so @rpath resolution works
    fix_dylib_install_name(&tmp);

    // Atomic rename into place
    if dest.exists() {
        fs::remove_dir_all(dest).ok();
    }
    fs::rename(&tmp, dest).expect("failed to move pdfium to cache dir");

    eprintln!("pdfium-sys: cached pdfium at {}", dest.display());
}

/// On macOS, pdfium-binaries ships dylibs with install name `./libpdfium.dylib`.
/// We need `@rpath/libpdfium.dylib` for rpath resolution to work.
fn fix_dylib_install_name(dir: &Path) {
    let target_os = env::var("CARGO_CFG_TARGET_OS").unwrap_or_default();
    if target_os != "macos" {

View on GitHub (pinned to 22d2dd8cd7)